/dingllm.nvim

Yacine's LLM nvim scripts

Primary LanguageLuaApache License 2.0Apache-2.0

wordart (6)

dingllm.nvim

Yacine's no frills LLM nvim scripts. free yourself, brothers and sisters

This is a really light config. I will be pushing breaking changes. I recommend reading the code and copying it over - it's really simple.

output.mp4

Credits

This extension woudln't exist if it weren't for https://github.com/melbaldove/llm.nvim

I diff'd on a fork of it until it was basically a rewrite. Thanks @melbaldove!

The main difference is that this uses events from plenary, rather than a timed async loop. I noticed that on some versions of nvim, melbaldove's extension would deadlock my editor. I suspected nio, so i just rewrote the extension.

lazy config

Add your API keys to your env (export it in zshrc or bashrc)

  {
    'yacineMTB/dingllm.nvim',
    dependencies = { 'nvim-lua/plenary.nvim' },
    config = function()
      local system_prompt =
        'You should replace the code that you are sent, only following the comments. Do not talk at all. Only output valid code. Do not provide any backticks that surround the code. Never ever output backticks like this ```. Any comment that is asking you for something should be removed after you satisfy them. Other comments should left alone. Do not output backticks'
      local helpful_prompt = 'You are a helpful assistant. What I have sent are my notes so far.'
      local dingllm = require 'dingllm'


      local function handle_open_router_spec_data(data_stream)
        local success, json = pcall(vim.json.decode, data_stream)
        if success then
          if json.choices and json.choices[1] and json.choices[1].text then
            local content = json.choices[1].text
            if content then
              dingllm.write_string_at_cursor(content)
            end
          end
        else
          print("non json " .. data_stream)
        end
      end

      local function custom_make_openai_spec_curl_args(opts, prompt)
        local url = opts.url
        local api_key = opts.api_key_name and os.getenv(opts.api_key_name)
        local data = {
          prompt = prompt,
          model = opts.model,
          temperature = 0.7,
          stream = true,
        }
        local args = { '-N', '-X', 'POST', '-H', 'Content-Type: application/json', '-d', vim.json.encode(data) }
        if api_key then
          table.insert(args, '-H')
          table.insert(args, 'Authorization: Bearer ' .. api_key)
        end
        table.insert(args, url)
        return args
      end


      local function llama_405b_base()
        dingllm.invoke_llm_and_stream_into_editor({
          url = 'https://openrouter.ai/api/v1/chat/completions',
          model = 'meta-llama/llama-3.1-405b',
          api_key_name = 'OPEN_ROUTER_API_KEY',
          max_tokens = '128',
          replace = false,
        }, custom_make_openai_spec_curl_args, handle_open_router_spec_data)
      end

      local function groq_replace()
        dingllm.invoke_llm_and_stream_into_editor({
          url = 'https://api.groq.com/openai/v1/chat/completions',
          model = 'llama-3.1-70b-versatile',
          api_key_name = 'GROQ_API_KEY',
          system_prompt = system_prompt,
          replace = true,
        }, dingllm.make_openai_spec_curl_args, dingllm.handle_openai_spec_data)
      end

      local function groq_help()
        dingllm.invoke_llm_and_stream_into_editor({
          url = 'https://api.groq.com/openai/v1/chat/completions',
          model = 'llama-3.1-70b-versatile',
          api_key_name = 'GROQ_API_KEY',
          system_prompt = helpful_prompt,
          replace = false,
        }, dingllm.make_openai_spec_curl_args, dingllm.handle_openai_spec_data)
      end

      local function llama405b_replace()
        dingllm.invoke_llm_and_stream_into_editor({
          url = 'https://api.lambdalabs.com/v1/chat/completions',
          model = 'hermes-3-llama-3.1-405b-fp8',
          api_key_name = 'LAMBDA_API_KEY',
          system_prompt = system_prompt,
          replace = true,
        }, dingllm.make_openai_spec_curl_args, dingllm.handle_openai_spec_data)
      end

      local function llama405b_help()
        dingllm.invoke_llm_and_stream_into_editor({
          url = 'https://api.lambdalabs.com/v1/chat/completions',
          model = 'hermes-3-llama-3.1-405b-fp8',
          api_key_name = 'LAMBDA_API_KEY',
          system_prompt = helpful_prompt,
          replace = false,
        }, dingllm.make_openai_spec_curl_args, dingllm.handle_openai_spec_data)
      end

      local function anthropic_help()
        dingllm.invoke_llm_and_stream_into_editor({
          url = 'https://api.anthropic.com/v1/messages',
          model = 'claude-3-5-sonnet-20240620',
          api_key_name = 'ANTHROPIC_API_KEY',
          system_prompt = helpful_prompt,
          replace = false,
        }, dingllm.make_anthropic_spec_curl_args, dingllm.handle_anthropic_spec_data)
      end

      local function anthropic_replace()
        dingllm.invoke_llm_and_stream_into_editor({
          url = 'https://api.anthropic.com/v1/messages',
          model = 'claude-3-5-sonnet-20240620',
          api_key_name = 'ANTHROPIC_API_KEY',
          system_prompt = system_prompt,
          replace = true,
        }, dingllm.make_anthropic_spec_curl_args, dingllm.handle_anthropic_spec_data)
      end

      vim.keymap.set({ 'n', 'v' }, '<leader>k', groq_replace, { desc = 'llm groq' })
      vim.keymap.set({ 'n', 'v' }, '<leader>K', groq_help, { desc = 'llm groq_help' })
      vim.keymap.set({ 'n', 'v' }, '<leader>L', llama405b_help, { desc = 'llm llama405b_help' })
      vim.keymap.set({ 'n', 'v' }, '<leader>l', llama405b_replace, { desc = 'llm llama405b_replace' })
      vim.keymap.set({ 'n', 'v' }, '<leader>I', anthropic_help, { desc = 'llm anthropic_help' })
      vim.keymap.set({ 'n', 'v' }, '<leader>i', anthropic_replace, { desc = 'llm anthropic' })
      vim.keymap.set({ 'n', 'v' }, '<leader>o', llama_405b_base, { desc = 'llama base' })
    end,
  },

Documentation

read the code dummy